Examples

dtfit examples

Self-contained, runnable scripts that walk the public API top to bottom. Each one is plain ASCII, prints its results (no plots required), and runs headless:

pip install -e packages/dtfit          # or: pip install dtfit
python packages/dtfit/examples/01_quickstart.py

Script	Covers
01_quickstart.py	first fit with fit(basis="legendre"), FittingResult (params / stderr / CI / prediction band), fit(basis="auto")
02_fitting_methods.py	the three batch fitters: fit(basis="legendre") (k_star, oscillatory recipe with freq_param), fit(basis="block") (robust=True, the block preset on a transient), fit_dsb (+ find_degree)
03_models_and_auto.py	the model catalog, self-seeding Model.fit, composition with +, suggest_models, fit(basis="auto") / auto_forecast
04_sklearn_estimator.py	NonlineRegressor with fit / predict / score, basis / order as estimator parameters, GridSearchCV, cross_val_score
05_streaming.py	ImageFilter(basis="block") (+ .tracking() / .robust() presets), ImageFilter(basis="legendre") (streaming amplitude + frequency), result(), pooled nis_ fused detection
06_scaling.py	fit_many (+ FittingProblem) fanned across CPU cores, ImageStream one-pass accumulation and map-reduce merge()
07_diagnostics.py	fit_report, residual_diagnostics, the image analytics, to_dict / from_dict, opt-in logging, and (with --plot) the *Display plots
08_stochastic.py	fit_stochastic / StochasticModel (regime, forecast bands, simulate round-trip), StochasticFilter online tracking, the Stochastic model wrapper

The only example with an optional dependency is 07_diagnostics.py --plot, which needs the viz extra (matplotlib): pip install "dtfit[viz]".

These scripts are executed in CI (tests/test_examples.py), so they stay in step with the API. The wiki renders them as the "Examples" guide pages.